On Fri, 24 Apr 1998, Brian Kelly wrote:
> Briefly the major development which should be of interest to members of
> this list is RDF, the Resource Description Framework. Tim Berners-Lee
> mentioned RDF several times in his keynote presentation (along with Dublin
> Core). RDF will not only provide a generic framwork for metadata, it will
> also provide "knowledge representation". This will enable more intelligent
> systems to be developed on the web.
To elaborate briefly on Brian's comments... RDF draws rather liberally on
the knowledge representation tradition in Artificial Intelligence. It
deliberately provides a rather simple framework, avoiding many of the
daunting complexities found in other languages like KIF or CycL. While RDF
is certainly not the last word in knowledge representation, it should
hopefully be enough to meet many of the requirements of the digital
libraries / metadata community. That said, it would be good to have more
detailed information about the specific requirements that eLib projects
might have...
Until the RDF schema language draft was published last week, there was
(due to the closed nature of W3C process) no easy way to gather
requirements from the eLib community. As a member of the RDF Schemas
working group, I found this pretty frustrating since RDF is extremely
relevant to many eLib projects. Now that we have published the
first (very preliminary) draft of the RDF Schema specification language,
it makes it a lot easier for me to ask for input and feedback from eLib
projects.
The RDF schema draft is available from http://www.w3.org/RDF/ and includes
an email address for submitting comments. If any eLib projects have any
specific concerns about the direction of RDF, I'd be happy to discuss them
off-list. My main concern right now is the RDF Schema language: is it too
simple? too complex? too confusing? We're trying to ensure that all
additional features added to the spec are grounded in real-world
requirements. With this in mind, input from eLib projects would be rather
valuable...
In the longer term, another issue worth bearing in mind is searchability:
it is likely that some mechanism for searching RDF Services will
eventually be defined. If so, what should we be doing now to plan for
this? To what extent should we be continue to assume that Z39.50 will be
the dominant search'n'retrieval protocol in 2-3 years time? What
requirements might we have of a "next generation" search protocol? What
features of Z39.50 (or WHOIS++) worked well or caused problems? (etc...)
Dan
ps. please direct technically oriented follups to lis-elib-tech or to
the (soon to be announced) rdf-dev list on mailbase.
--
[log in to unmask]
Research and Development Unit tel: +44(0)117 9288478
Institute for Learning and Research Technology http://www.ilrt.bris.ac.uk/
University of Bristol, Bristol BS8 1TN, UK. fax: +44(0)117 9288473
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|